“The AI Vulnerability Storm: Building a Mythos-Ready Security Program” delivers a risk register, 11 priority actions, and board briefing framework built by 60+ contributors and reviewed by 250+ CISOs in a single weekend

April 14, 2026. SANS Institute and the Cloud Security Alliance (CSA), alongside [un]prompted and the OWASP GenAI Security Project, today released “The AI Vulnerability Storm: Building a Mythos-Ready Security Program,” a free strategy briefing that gives CISOs and security leaders an actionable framework for responding to the accelerating pace of AI-driven vulnerability discovery and exploitation.

The briefing was produced over a single weekend by more than 60 named contributors and reviewed by over 250 CISOs from across the global cybersecurity community. It responds directly to the capabilities demonstrated by Anthropic’s Claude Mythos (Preview) and Project Glasswing, which autonomously identified thousands of zero-day vulnerabilities across every major operating system and web browser, including a 27-year-old vulnerability in OpenBSD, one of the most security-hardened operating systems in the world.

“The window between vulnerability discovery and weaponization has collapsed into hours,” said Rob T. Lee, Chief AI Officer and Chief of Research at SANS Institute and co-author of the briefing. “What Mythos shows us is a permanent acceleration. This document gives CISOs something the commentary doesn’t: a risk register, priority actions with start dates, and a board briefing they can use this week.”

A 12-MONTH ESCALATION

The briefing documents a rapid escalation in AI offensive capabilities over the past year. In June 2025, XBOW became the first autonomous system to top HackerOne’s US leaderboard, outperforming all human hackers on the platform. In August 2025, DARPA’s AI Cyber Challenge found 54 vulnerabilities across 54 million lines of code in four hours. By November 2025, Anthropic disclosed a Chinese state-sponsored group had used AI to autonomously run full attack chains, from reconnaissance through data exfiltration, across approximately 30 global targets.

In February 2026, Anthropic reported more than 500 high-severity vulnerabilities in open source software using Claude Opus 4.6. Sysdig documented an AI-based attack that reached administrator-level access in eight minutes. Linux kernel maintainers saw vulnerability reports climb from two to ten per week.

Mythos represents a further step change. In internal testing, the model generated 181 working exploits against Firefox vulnerabilities where the previous best model succeeded only twice under the same conditions. The model achieved a 72% exploit success rate and demonstrated the ability to chain multiple vulnerabilities into single exploit paths without human guidance.

According to the Zero Day Clock, the mean time from vulnerability disclosure to confirmed exploitation has fallen to less than one day in 2026, down from 2.3 years in 2019.

WHAT THE BRIEFING INCLUDES

The briefing includes a 13-item risk register mapped to four industry frameworks (OWASP LLM Top 10 2025, OWASP Agentic Top 10 2026, MITRE ATLAS, and NIST CSF 2.0), an 11-item priority actions table with aggressive timelines, 10 diagnostic questions for CISOs to triage their current security program, and a board-ready executive briefing section.

Key findings:

AI-driven vulnerability discovery tools can now generate working exploits at a rate that outpaces organizational patch cycles. Every patch also becomes an exploit blueprint, as AI accelerates patch-diffing and reverse engineering of fixes.

Defensive teams that have not adopted AI agents face a widening capability gap against AI-augmented adversaries, regardless of their existing technical skill. The briefing classifies this as a cultural challenge as much as a technological one.

The EU AI Act takes effect in August 2026, introducing automated audit, incident reporting, and cybersecurity requirements around AI. When AI can find vulnerabilities at accessible cost, the standard for what constitutes reasonable defensive effort shifts, creating direct governance and liability exposure for organizations that do not adapt.

Organizations should prepare for a sustained increase in the volume and cadence of vulnerability disclosures and should plan for operational burnout as security teams absorb the increase without corresponding investment in headcount or tooling.

The briefing’s first priority action skips governance entirely: point AI agents at your own code this week. The longest-horizon item is to stand up a permanent Vulnerability Operations (VulnOps) function within 12 months, staffed and automated for continuous AI-driven discovery across the entire software estate.

“Attackers already operate as syndicates, crowdsourcing, sharing tools, moving as a collective. Defenders have to do the same,” said Gadi Evron, CEO of Knostic and CISO-in-Residence for AI at the Cloud Security Alliance, and lead author of the briefing. “We built this in three days because CISOs needed it now, not when it was perfect. Mythos is the first wave. The organizations that build the muscle now (the processes, the tooling, and a culture willing to adopt AI as a core part of how security gets done) will be the ones that meet the next wave on their own terms.”

Download the paper now.

CONTRIBUTING AUTHORS AND REVIEWERS

The briefing was co-authored by Gadi Evron (CEO, Knostic; CISO-in-Residence for AI, Cloud Security Alliance), Rob T. Lee (Chief AI Officer and Chief of Research, SANS Institute), and Rich Mogull (Chief Analyst, Cloud Security Alliance).

Contributing authors include Jen Easterly (CEO, RSAC; former Director, CISA), Bruce Schneier (Chief of Security Architecture, Inrupt; Fellow and Lecturer, Harvard Kennedy School), Chris Inglis (former National Cyber Director, The White House), Phil Venables (Ballistic Ventures; former CISO, Google Cloud), Heather Adkins (CISO, Google), Rob Joyce (former Cybersecurity Director, NSA), Sounil Yu (CTO, Knostic; former Chief Security Scientist, Bank of America), Katie Moussouris (Founder and CEO, Luta Security), James Lyne (CEO, SANS Institute), Jim Reavis (CEO, Cloud Security Alliance), Joshua Saxe (CTO, Security Superintelligence Labs; former AI and Llama Security Lead, Meta), and others.

Reviewers include CISOs from Atlassian, Cloudflare, Cushman & Wakefield, GitLab, Google, Hyland, NFL, Netflix, Rivian, Sophos, TransUnion, Wells Fargo, and more than 40 additional organizations. SANS reviewers include Chris Cochran (Field CISO and VP of AI Security), Ed Skoudis (President, SANS Technology Institute), and Ciaran Martin (Head of Cyber Leaders Network; Founder and former CEO, UK NCSC).

SANS CRITICAL ADVISORY: BUGBUSTERS - AI VULNERABILITY DISCOVERY HYPE VS. REALITY

Live Webcast | Thursday, April 16

SANS Institute will host SANS Critical Advisory: BugBusters - AI Vulnerability Discovery Hype vs. Reality, a live demonstration of AI-assisted vulnerability discovery against real code, on Thursday, April 16 at 12:00 PM Noon ET. SANS Technology Institute President Ed Skoudis, Faculty Fellow Joshua Wright, and Principal Instructor Chris Elgee will demonstrate the practical application of current AI models to vulnerability research, based on 15 months of real-world penetration testing experience. No registration is required.

Tune in to watch live on April 16.

SANS AI CYBERSECURITY SUMMIT 2026

In-Person in Arlington, VA & Free Online | April 20–21

The SANS AI Cybersecurity Summit brings together practitioners, researchers, and security leaders for two days of hands-on sessions on defending against AI-powered threats and deploying AI securely within security operations. Speakers include contributors to the Mythos-Ready briefing alongside leading voices in AI security from across the industry. The summit is free to attend online.

Register for the SANS AI Cybersecurity Summit.

CSA AGENTIC AI SECURITY SUMMIT 2026

Virtual Event | April 29-30

CSA's Agentic AI Security Summit is a free, two-day virtual event dedicated to helping organizations securely design, deploy, and scale autonomous and collaborative AI systems. Across two immersive days, global leaders in AI governance, cybersecurity, cloud infrastructure, and standards development will share practical strategies for securing the next generation of AI ecosystems. From emerging risks to real-world implementation guidance, this summit delivers the clarity and actionable insight security and AI leaders need now.

Register now to help shape the future of secure Agentic AI.

ABOUT THE BRIEFING

“The AI Vulnerability Storm: Building a Mythos-Ready Security Program” is available free at: https://labs.cloudsecurityalliance.org/mythos-ciso/

Audio and video overviews are available at: https://sansurl.com/CISO-mythos-ready

The briefing is released under the Attribution-NonCommercial 4.0 International (CC BY-NC 4.0) license. All listed authors and reviewers represent only themselves.

ABOUT SANS INSTITUTE
The SANS Institute was established in 1989 as a cooperative research and education organization. Today, SANS is the most trusted and, by far, the largest provider of cybersecurity training and certification to professionals in government and commercial institutions worldwide. Renowned SANS instructors teach more than 85 courses at in-person and virtual cybersecurity events and OnDemand. GIAC, an affiliate of the SANS Institute, validates practitioner skills through more than 50 hands-on technical certifications in cybersecurity.

ABOUT CLOUD SECURITY ALLIANCE
The Cloud Security Alliance (CSA) is the world’s leading not-for-profit organization committed to awareness, practical implementation, and credentialing of forward-looking cybersecurity topics, including AI, cloud, and Zero Trust. In an era where digital transformation drives business success, CSA stands as the global authority ensuring organizations can operate securely while harnessing cutting-edge technology. Through volunteer-driven research, globally-accepted standards, the CSAI Foundation, and award-winning vendor-neutral education programs that unite technical experts, industry practitioners, and varied associations, governments, chapters, and corporate members, CSA bridges the gap between innovation and pragmatic security execution. Visit CSA’s website and the CSAI Foundation, a 501(c)3 non-profit foundation dedicated exclusively to AI security and safety, to learn more.    

###